The debunking of superluminal neutrinos

The July issue of Significance includes an article in its "Bad Stats" series by Robert Matthews about the claim made in 2011 by experimental physicists in San Grasso, Italy, of detecting faster-than-light neutrinos from CERN.  This was the story of the OPERA experiment (Oscillation Project with Emulsion-tRacking Apparatus).  Among the literature cited by Matthews is a very good write-up on the episode by Chad Orzel, in the MIT Ptess Reader.  This article originally appeared in a 2020 book, Pseudoscience, published by MIT Press.  Orzel's piece is probably the best non-technical summary and post-mortem of the incident that I've run across - a keeper.  DTLR recommends it!

Happy Birthday, APS!

May 20 was the 125th birthday of the American Physical Society (APS).  Founded in 1899 at Columbia University, and later taking over publication of the Physical Review (founded in 1893 at Cornell University) in 1913, the society has expanded both its core early functions (conferences and journal publications) as well as added many other programs in the last century and a quarter.

I joined the APS around 1997, and its Centennial Meeting in Atlanta (1999) was the first national scientific conference I ever attended, and the first one I ever gave a talk at.  That year, the March and April meetings were combined into a grand extravaganza, with (if I recall right) well over 10,000 attendees.  I remember following (by coincidence) at least one Nobel Laureate down an escalator (J. Friedman) and recognizing others in the hallways or meeting rooms.  I presented again at the 2002 March Meeting in Indianapolis, though my findings were shot down by a very prominent scientist in the audience.  That was the second and final time I presented at an APS conference, though in later years I've attended "for fun" on my own time on several occasions, including last year.  I have also attended a number of APS section meetings, as well as the Division of Fluid Dynamics annual meetings.  I once published in an APS journal (Physical Review E), but that was during the end of my time in graduate school.

Since I have not been a practicing physicist for over two decades, I have not been an active APS member, but I am a Life Member of the parent organization and a number of its individual units.  Indeed, I have been an APS member for just over 20% of APS's own lifetime, and just over 50% of my own lifetime!  Many Happy Returns, APS!

Physics Today and fluid mechanics

Back in December, I discussed a couple recent items in Physics Today of great relevance to fluid mechanics.  I remarked that we don't get to see fluid mechanics on the cover of PT very often.  Well, lo and behold, the last couple months have proven me delightfully wrong!  The cover of the April issue featured "Fluid Dynamics of Dry Salt Lakes", referring to a "quick study" article by Beaume, Goehring, and Lasser.  It discusses a theory of groundwater convection as a possible explanation of polygonal patterns on dry lake beds ("salt polygons").  The very next month, the cover features "A Shocking Start to Stars", referring to a piece by Ceccarelli and Codella, on the role of shock waves in the interstallar medium, especially in star formation, and even in synthesizing certain precursor molecules of life itself.

However, I was personally most enthralled by another piece that also appeared in the May issue by meteorologist Tim Palmer, "The real butterfly effect and maggoty applies".  The title seems to have been taken from two earlier efforts, one coauthored by Palmer in 2019 on the real butterfly effect (published in Nonlinearity), and the other a 2002 PT Reference Frame article by Sir Michael Berry on singular limits.  Both these papers seem to have had a strong influence on this article, which mainly deals with weather predictibility, and more generally the predictability of nonlinear systems, including the limits of artificial intelligence-based prediction.  The article is very good, informative, and very readable.  I'd come across the singular limit concept in discussions of aerodynamic lift, where viscosity is the parameter that when taken to exactly zero (and not just a limit approaching zero) results in the impossibility of lift.  Reading Palmer's piece, and then Berry's, has helped me understand that there are other such examples in physics.

Another howler from Science Magazine

In my last post, I scolded the editor-in-chief of Science, Holden Thorp, for an op-ed, though I largely agreed with his conclusion.  It didn't take long for him to provoke the DTLR hornet's nest again, and this time I can't say I agree with him.  The April 11 issue features another op-ed by him, titled "Teach Philosophy of Science".  He points out the risk of the erosion of public trust in science, pointing specifically to survey results indicating that 92% of Americans "felt it important that scientists show they are 'open to changing their minds based on new evidence,' which is of course what they must do," Thorp writes.  He claims that the "history of science is a powerful narrative of this culture of self-correction" and laments that "Resetting the public's understanding of how science works will be a big job."  His proposed solution is to teach philosophy of science to undergraduate science majors!

I regard that proposed solution as a non-sequitur.  Using Thorp's own argument, I say it is the history, not the philosophy of science that should be taught.  To be honest though, when I was a young student, I was interested in history, but not the history of science.  I was busy learning the science itself.  I am much more receptive to history of science now, for the reasons Thorp suggests.  I have not found philosophy of science to be insightful or to help me think about doing science.  Earlier this year I finished reading A Philospher Looks at Science, by Nancy Cartwright.  While I admire her support for the views of Kay and King's Radical Uncertainty, her book hardly made a dent on my way of thinking about science, and how to do science, whereas Kay and King helped me substantially clarify my thinking on probability models applied to the real world.

Returning to Thorp, the bigger problem in his argument is that somehow the public will trust science more if only they understood better how scientists change their views and "self-correct".  Unfortunately the history of science has plenty of examples of abject failures as well as successes.  Consider the wholly unnecesssary resistance to such figures as Ignaz Semmelweiss (who contributed to the germ theory of disease), Dan Shechtman (discoverer of quasicrystals), and Katalin Kariko (mRNA pioneer).  Consider also the crisis of non-reproducible research that I mentioned in my last post.  Science won't self-correct if no one bothers doing the self-correcting.  There are no financial or professional incentives either to reproduce others' work, nor to do one's own work with sufficient care and sound methodology to ensure it will withstand the test of time.

At its most cynical, science is merely a self-propagating exercise in securing endless piles of grant money for "research" of limited value and interest, as noted in this recent video by Sabine Hossenfelder, who points out that it is public tax money that fuels the seemingly joyless cycle of beg, spend, and publish.  If the public saw science this way, rather than in the romanticized way Thorp imagines, would they really trust scientists enough to allow their taxes to be spent as such?  Personally I would not, and I am a scientist!

I assert that the number one thing scientists can do to increase public trust in science is to reform their own house.  Only then will they deserve public support and trust.

Digression.  I last wrote about Sabine Hossenfelder nine years ago this month.  Although in that post I largely disagreed with her on one particular topic, I've since come to admire her devastingly blunt contrarian views on physics research.  For some time, I was also a financial supporter of hers on Patreon, though I am not longer financially able to support any of the deserving people I admire on Patreon.

Science Magazine colossal hypocrisy

I could not believe my eyes when I opened the Feb. 2 issue of Science to the editorial page.  The featured op-ed, as frequently is the case, is by the current editor-in-chief, Holden Thorp, and is entitled "Earning Respect and Trust".  He discussed the Science family of journals' editorial staffs, and the lack of respect they sometimes encounter by authors of submitted manuscripts.  He rightfully defends these staffers, and insists that they be respected by lab scientists.  I wholly agree with these sentiments.  It is his method of argument that I strongly reject.

Let's look, for example, at a few sentences in the third paragraph, line by line.  First,

If anything, the challenges that science is experiencing now are not due to a lack of success in the laboratory.

That is manifestly false.  The crisis of non-reproducible research, documented for nearly 20 years now, with almost no widespread progress on reform, has been a frequent topic of discussion on DTLR, and even in the news section of Science itself.  Two of the cornerstone enablers of this crisis have been poor methodology, both in the lab and in the data analysis.  I have specifically called out previous AAAS CEO Rush Holt's similar denials that it is exactly the core activity of scientific research that is rotten; for example:  here, here, here, and here.  I won't rehash those arguments.  Holt's retirement from AAAS, the year after my posts, was greatly welcomed.  However. Holden Thorp has now revived what I will call the Holt Myth.

Next,

The notion that lab work is the only purposeful endeavor in science is obtuse and is an example of precisely what leads to the view that scientists are intellectual elites who do not value the contributions or abilities of anyone except themselves and the small group they deign to recognize as their peers.  Every time this academic hauteur is revealed to the public, confidence is lost for a simple reason--scientists like these are not inspiring the people's trust.

Now, I completely agree with this criticism!  However, he uses the term "top journal" earlier in the paragraph, not quite saying, but seemingly implying, that his journal is among the top ones.  In many fields Science certainly is considered among the top journals (e.g., see here), and he is part of the very elite he seems to be complaining about.  Authors who aim to publish in such elite journals, where novelty is prized over reproducibility, are incentivized to be sloppy and fast, rather than careful and slow.  The motto of DTLR is "Garbage In, Gospel Out":  Thorp's journal could be exhibit A of this phenomenon.  Perhaps Science and its editors should themselves take a humility pill before lecturing the rest of us on the harms of elitism.

 

 

Aerodynamic lift

It is natural to expect anyone who claims to be a physicist to be able to give a simple, intuitive explanation of aerodynamic lift, using basic principles of classical mechanics.  Unfortunately, most of the time a physicist's such explanation is misleading, over-simplified, or just plain wrong.  One of the most important reasons for this is that there is no such simple, intuitive explanation.  I recently stumbled upon a reference to this excellent 2020 Scientific American article by Ed Regis, that outlines the two main competing explanations (Bernoulli's theorem vs. Newton's third law).  However, others such as the Coanda effect, have been proposed, and are not mentioned in the article.  Nonetheless Regis' article resonated with me because he seems to give a balanced discussion while debunking some of the clearly incorrect variants of these arguments.  He also spoke with two people whose insights I regard as essential, namely, John D. Anderson and Doug McLean.  Anderson is the author of a number of widely used textbooks on relevant topics, such as Introduction to Flight, Fundamentals of Aerodynamics, Modern Compressible Flow, Computational Fluid Dynamics: The Basics with Applications, and A History of Aerodynamics.  (I own three of these books.)  McLean is a retired Boeing engineer, and author of Understanding Aerodynamics, which I recently acquired.

I became interested in the competing intuitive explanations for aerodynamic lift during my time in graduate school, in the late 1990s.  At the time I was involved in atmospheric science research.  One of the professors I worked with informed the members of his lab that Prof. Anderson had been invited to visit our university; and moreover (knowing of my interest in this topic) my professor had invited Anderson to visit our lab for a discussion on it.  It was certainly an honor, and I and at least one other student in the group asked Prof. Anderson to sign our textbooks during his visit.  At the time, I was very partial to the Newton's third law version of the explanation (and still am, though I recognize it as woefully incomplete).  Anderson struck a more neutral tone, stating that both the Bernoulli principle and Newton's second law were at work in lifting an airfoil.  Neither of them was "wrong" per se.  That discussion over 20 years ago seems compatible with the views he gives in Regis' piece.

I won't discuss the officially accepted explanations, formulated in the early 20th century, given by Kutta and Zhukovsky, for 2-dimensional flow, and by Prandtl, Lanchester, and Prandtl's students (notably Blasius) for 3-dimensional flow, here.  However I note that Falkovich (Fluid Mechanics, 2/e, Cambridge University Press, 2018) points to the importance of viscosity in these formulations.  A purely inviscid flow "produces no lift" (p. 56).  He continues, "Without friction-caused separation [of the boundary layer], birds and planes would not be able to fly."  Neither the Bernoulli principle (which is strictly only true for inviscid flows) nor the Newton's third law explanation makes any reference to boundary layer separation and nonzero circulation around the airfoil.  As Anderson says in the Regis article, there just isn't a "one-liner" explanation of aerodynamic lift.

Galileo's Assayer and Torricelli's barometer

A couple months ago, Science magazine had a retrospective by Alex Gomez-Marin on Galileo's The Assayer, which was published 400 years ago this year.  Part of a series of scientific disputes regarding the nature of comets, this particular piece lays out Galileo's methodological perspective on how science should be done.  I've only read selected excerpts in translation, but I do know that others (e.g., Flemish contemporary Simon Stevin) held similar views.

Meanwhile the December issue of Physics Today features a Backscatter item on Torricelli's barometer, by Karina Cooper.  Torricelli was Galileo's last "student", and his work with the barometer was an important step in the understanding of hydrostatic pressure.  Together with Torricelli's efflux law, this achievement ranks him as a major founder of 17th century fluid mechanics, along with Stevin and Galileo's older student, Benedetto Castelli.

My interest in these matters, and the history of classical physics more generaly, has grown noticeably in the last year or so, particularly in the arena of fluid mechanics, which is an interdisciplinary field involving physics, mathematics, and engineering.  Seeing these items in Science and Physics Today is a great way to wind down the year of 2023.  

 As a bonus I praise Physics Today for featuring a cover story in November on research in contemporary fluid mechanics, namely the piece by Detlef Lohse and Olga Shishkina on "Ultimate turbulent thermal convection".  However, the authors' insistence on the relevance of the reported work for real world situations, such as atmospheric and photospheric convection, seems to me insincere when they focus on results for "classical regime" aspect ratios between 0.5 and 1.0.  The atmospheric and photospheric examples involve large aspect ratios.

Nonetheless, the cover is worth reproducing here.  We don't get to see fluid mechanics featured on the front cover of Physics Today all that often!

If I don't write again this year, may you all have a wonderful holiday season and a blessed new year.  See you in 2024!